maintenance/orphans.php

   1 <?php
   2 /**
   3  * Look for 'orphan' revisions hooked to pages which don't exist and
   4  * 'childless' pages with no revisions.
   5  * Then, kill the poor widows and orphans.
   6  * Man this is depressing.
   7  *
   8  * Copyright © 2005 Brion Vibber <brion@pobox.com>
   9  * https://www.mediawiki.org/
  10  *
  11  * This program is free software; you can redistribute it and/or modify
  12  * it under the terms of the GNU General Public License as published by
  13  * the Free Software Foundation; either version 2 of the License, or
  14  * (at your option) any later version.
  15  *
  16  * This program is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  19  * GNU General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License along
  22  * with this program; if not, write to the Free Software Foundation, Inc.,
  23  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  24  * http://www.gnu.org/copyleft/gpl.html
  25  *
  26  * @file
  27  * @author <brion@pobox.com>
  28  * @ingroup Maintenance
  29  */
  30
  31 require_once __DIR__ . '/Maintenance.php';
  32
  33 use Wikimedia\Rdbms\IMaintainableDatabase;
  34
  35 /**
  36  * Maintenance script that looks for 'orphan' revisions hooked to pages which
  37  * don't exist and 'childless' pages with no revisions.
  38  *
  39  * @ingroup Maintenance
  40  */
  41 class Orphans extends Maintenance {
  42         public function __construct() {
  43                 parent::__construct();
  44                 $this->addDescription( "Look for 'orphan' revisions hooked to pages which don't exist\n" .
  45                         "and 'childless' pages with no revisions\n" .
  46                         "Then, kill the poor widows and orphans\n" .
  47                         "Man this is depressing"
  48                 );
  49                 $this->addOption( 'fix', 'Actually fix broken entries' );
  50         }
  51
  52         public function execute() {
  53                 $this->checkOrphans( $this->hasOption( 'fix' ) );
  54                 $this->checkSeparation( $this->hasOption( 'fix' ) );
  55                 # Does not work yet, do not use
  56                 # $this->checkWidows( $this->hasOption( 'fix' ) );
  57         }
  58
  59         /**
  60          * Lock the appropriate tables for the script
  61          * @param IMaintainableDatabase $db
  62          * @param string[] $extraTable The name of any extra tables to lock (eg: text)
  63          */
  64         private function lockTables( $db, $extraTable = [] ) {
  65                 $tbls = [ 'page', 'revision', 'redirect' ];
  66                 if ( $extraTable ) {
  67                         $tbls = array_merge( $tbls, $extraTable );
  68                 }
  69                 $db->lockTables( [], $tbls, __METHOD__, false );
  70         }
  71
  72         /**
  73          * Check for orphan revisions
  74          * @param bool $fix Whether to fix broken revisions when found
  75          */
  76         private function checkOrphans( $fix ) {
  77                 $dbw = $this->getDB( DB_MASTER );
  78                 $commentStore = new CommentStore( 'rev_comment' );
  79
  80                 if ( $fix ) {
  81                         $this->lockTables( $dbw );
  82                 }
  83
  84                 $commentQuery = $commentStore->getJoin();
  85
  86                 $this->output( "Checking for orphan revision table entries... "
  87                         . "(this may take a while on a large wiki)\n" );
  88                 $result = $dbw->select(
  89                         [ 'revision', 'page' ] + $commentQuery['tables'],
  90                         [ 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text' ] + $commentQuery['fields'],
  91                         [ 'page_id' => null ],
  92                         __METHOD__,
  93                         [],
  94                         [ 'page' => [ 'LEFT JOIN', [ 'rev_page=page_id' ] ] ] + $commentQuery['joins']
  95                 );
  96                 $orphans = $result->numRows();
  97                 if ( $orphans > 0 ) {
  98                         global $wgContLang;
  99
 100                         $this->output( "$orphans orphan revisions...\n" );
 101                         $this->output( sprintf(
 102                                 "%10s %10s %14s %20s %s\n",
 103                                 'rev_id', 'rev_page', 'rev_timestamp', 'rev_user_text', 'rev_comment'
 104                         ) );
 105
 106                         foreach ( $result as $row ) {
 107                                 $comment = $commentStore->getComment( $row )->text;
 108                                 if ( $comment !== '' ) {
 109                                         $comment = '(' . $wgContLang->truncate( $comment, 40 ) . ')';
 110                                 }
 111                                 $this->output( sprintf( "%10d %10d %14s %20s %s\n",
 112                                         $row->rev_id,
 113                                         $row->rev_page,
 114                                         $row->rev_timestamp,
 115                                         $wgContLang->truncate( $row->rev_user_text, 17 ),
 116                                         $comment ) );
 117                                 if ( $fix ) {
 118                                         $dbw->delete( 'revision', [ 'rev_id' => $row->rev_id ] );
 119                                 }
 120                         }
 121                         if ( !$fix ) {
 122                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 123                         }
 124                 } else {
 125                         $this->output( "No orphans! Yay!\n" );
 126                 }
 127
 128                 if ( $fix ) {
 129                         $dbw->unlockTables( __METHOD__ );
 130                 }
 131         }
 132
 133         /**
 134          * @param bool $fix
 135          * @todo DON'T USE THIS YET! It will remove entries which have children,
 136          *       but which aren't properly attached (eg if page_latest is bogus
 137          *       but valid revisions do exist)
 138          */
 139         private function checkWidows( $fix ) {
 140                 $dbw = $this->getDB( DB_MASTER );
 141                 $page = $dbw->tableName( 'page' );
 142                 $revision = $dbw->tableName( 'revision' );
 143
 144                 if ( $fix ) {
 145                         $this->lockTables( $dbw );
 146                 }
 147
 148                 $this->output( "\nChecking for childless page table entries... "
 149                         . "(this may take a while on a large wiki)\n" );
 150                 $result = $dbw->query( "
 151                         SELECT *
 152                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 153                         WHERE rev_id IS NULL
 154                 " );
 155                 $widows = $result->numRows();
 156                 if ( $widows > 0 ) {
 157                         $this->output( "$widows childless pages...\n" );
 158                         $this->output( sprintf( "%10s %11s %2s %s\n", 'page_id', 'page_latest', 'ns', 'page_title' ) );
 159                         foreach ( $result as $row ) {
 160                                 printf( "%10d %11d %2d %s\n",
 161                                         $row->page_id,
 162                                         $row->page_latest,
 163                                         $row->page_namespace,
 164                                         $row->page_title );
 165                                 if ( $fix ) {
 166                                         $dbw->delete( 'page', [ 'page_id' => $row->page_id ] );
 167                                 }
 168                         }
 169                         if ( !$fix ) {
 170                                 $this->output( "Run again with --fix to remove these entries automatically.\n" );
 171                         }
 172                 } else {
 173                         $this->output( "No childless pages! Yay!\n" );
 174                 }
 175
 176                 if ( $fix ) {
 177                         $dbw->unlockTables( __METHOD__ );
 178                 }
 179         }
 180
 181         /**
 182          * Check for pages where page_latest is wrong
 183          * @param bool $fix Whether to fix broken entries
 184          */
 185         private function checkSeparation( $fix ) {
 186                 $dbw = $this->getDB( DB_MASTER );
 187                 $page = $dbw->tableName( 'page' );
 188                 $revision = $dbw->tableName( 'revision' );
 189
 190                 if ( $fix ) {
 191                         $this->lockTables( $dbw, [ 'user', 'text' ] );
 192                 }
 193
 194                 $this->output( "\nChecking for pages whose page_latest links are incorrect... "
 195                         . "(this may take a while on a large wiki)\n" );
 196                 $result = $dbw->query( "
 197                         SELECT *
 198                         FROM $page LEFT OUTER JOIN $revision ON page_latest=rev_id
 199                 " );
 200                 $found = 0;
 201                 foreach ( $result as $row ) {
 202                         $result2 = $dbw->query( "
 203                                 SELECT MAX(rev_timestamp) as max_timestamp
 204                                 FROM $revision
 205                                 WHERE rev_page=$row->page_id
 206                         " );
 207                         $row2 = $dbw->fetchObject( $result2 );
 208                         if ( $row2 ) {
 209                                 if ( $row->rev_timestamp != $row2->max_timestamp ) {
 210                                         if ( $found == 0 ) {
 211                                                 $this->output( sprintf( "%10s %10s %14s %14s\n",
 212                                                         'page_id', 'rev_id', 'timestamp', 'max timestamp' ) );
 213                                         }
 214                                         ++$found;
 215                                         $this->output( sprintf( "%10d %10d %14s %14s\n",
 216                                                 $row->page_id,
 217                                                 $row->page_latest,
 218                                                 $row->rev_timestamp,
 219                                                 $row2->max_timestamp ) );
 220                                         if ( $fix ) {
 221                                                 # ...
 222                                                 $maxId = $dbw->selectField(
 223                                                         'revision',
 224                                                         'rev_id',
 225                                                         [
 226                                                                 'rev_page' => $row->page_id,
 227                                                                 'rev_timestamp' => $row2->max_timestamp ] );
 228                                                 $this->output( "... updating to revision $maxId\n" );
 229                                                 $maxRev = Revision::newFromId( $maxId );
 230                                                 $title = Title::makeTitle( $row->page_namespace, $row->page_title );
 231                                                 $article = WikiPage::factory( $title );
 232                                                 $article->updateRevisionOn( $dbw, $maxRev );
 233                                         }
 234                                 }
 235                         } else {
 236                                 $this->output( "wtf\n" );
 237                         }
 238                 }
 239
 240                 if ( $found ) {
 241                         $this->output( "Found $found pages with incorrect latest revision.\n" );
 242                 } else {
 243                         $this->output( "No pages with incorrect latest revision. Yay!\n" );
 244                 }
 245                 if ( !$fix && $found > 0 ) {
 246                         $this->output( "Run again with --fix to remove these entries automatically.\n" );
 247                 }
 248
 249                 if ( $fix ) {
 250                         $dbw->unlockTables( __METHOD__ );
 251                 }
 252         }
 253 }
 254
 255 $maintClass = "Orphans";
 256 require_once RUN_MAINTENANCE_IF_MAIN;